The U.S. Geological Survey (USGS) has undertaken a mineral resources assessment for tungsten for a portion of the Great Basin in parts of western Nevada and east-central California. This data release provides the Great Basin Tungsten Database: the geospatial and geologic data, and results of chemical analyses for 46,955 samples collected in the assessment area, extracted from the USGS National Geochemical Database. These rock records were collected as part of various programs and projects at the USGS and analyzed from 1963 to 2015. The database represents rock records, each comprising one best value chemical determination for each analyzed chemical species, that include skarns, carbonate lithologies (for example, limestones, dolostones, siliclastic-carbonates, metacarbonates), and granitoid lithologies (intrusive, hypabyssal, and extrusive rocks containing greater than or equal to 65 weight percent SiO2). It was compiled to integrate geochemical rock data along with geologic mapping, mineral sites, geophysical, remote sensing, and watershed data to facilitate the assessment of tungsten skarn mineral resource potential in the region. In general, mineral resources assessments rely heavily on already available databases of mineral sites descriptions and geochemical results of analyses carried out on samples collected over the years with different exploration objectives. As is the case with many large databases that have been built over the course of many years and are therefore the result of contributions from many people and projects with specific and independent objectives, there might be a lack of consistency of the data included and some records might be richly attributed, with complete descriptions while others might be unusable due to the lack of criteria to consider them reliable. To deal with this situation, this subset has been revised, refined, and improved for usability in the assessment and the results presented here are what is considered as “best values” of those analyses. The methodology employed to obtain “best values” from analytical data, described by Granitto and others (2019), takes into consideration the variables that intervened at the time of the analysis, such as sample weight, sample decomposition or digestion method, analytical instrumentation, sensitivity, reliability, age of the method, limits of detection; and produces a series of tables that rank, from best to least preferred, the analytical methods that produced the values reported for each element in the database. The data provided in this compilation are the most comprehensive and accurate to date and should be useful for various geochemical studies across this region.